Wide Plate Margin Deformation, Southern Central America and Northwestern South America, CASA GPS Observations

Home , Caribbean Plate, South American Plate

Journal of South American Earth Sciences 15 62002) 157±171 www.elsevier.com/locate/jsames

Wide plate margin deformation, southern Central America and northwestern South America, CASA GPS observations

Robert Trenkampa,*, James N. Kellogga, Jeffrey T. Freymuellerb, Hector P. Morac

aDepartment of Geological Sciences, University of South Carolina, Columbia, SC 29208, USA bGeophysical Institute, University of Alaska, Fairbanks, AK 99775, USA cVolcanological and Seismological Observatory, INGEOMINAS, Avenida 12 de Octubre 15-47, Manizales, Colombia Received 1 November 2000; accepted 1 January 2002

Abstract Global positioning system data from southern Central America and northwestern South America were collected during 1991, 1994, 1996, and 1998 in Costa Rica, Panama, Ecuador, Colombia, and Venezuela. These data reveal wide plate boundary deformation and escape tectonics occurring along an approximately 1400 km length of the North Andes, locking of the subducting Nazca plate and strain accumulation in the Ecuador±Colombia forearc, ongoing collision of the Panama arc and Colombia, and convergence of the Caribbean plate with Panama and South America. Elastic modeling of observed horizontal displacements in the Ecuador forearc is consistent with partial locking 650%) in the subduction zone and partial transfer of motion to the overriding South American plate. The deformation is hypothesized to re¯ect elastic recoverable strain accumulation associated with the historic seismicity of the area and active faulting associated with permanent shortening of 6 mm/a. Deformation associated with the Panama±Colombia collision is consistent with elastic strain accumulation on a fully locked Atrato±Uraba Fault Zone suture. q 2002 Elsevier Science Ltd. All rights reserved.

Keywords: GPS; North Andes; oblique subduction; arc continent collision; earthquake strain; escape; caribbean

1. Introduction The Central and South America 6CASA) GPS project was inaugurated in 1988 to study plate motions and crustal Mid-ocean ridge and transform fault azimuths, spreading deformation in a tectonically active area of complex inter- rates, and earthquake slip vectors at plate boundaries have action among the Nazca, Cocos, Caribbean, and South successfully explained large-scale features of plate kine- American plates 6Fig. 1). The tectonic development of the matics and demonstrated that plate interiors generally CASA area has been the object of multiple geologic and behave rigidly over geologic time scales 6Le Pichon, geophysical studies 6for a comprehensive listing, see 1968; Minster et al., 1974; Minster and Jordan, 1978; Kellogg and Vega, 1995; Ego et al., 1996; Gutscher et al., Chase, 1978; DeMets et al., 1990). However, convergent 1999 and references therein). Previous studies have shown plate boundaries were recognized early in the development the CASA region to be a complex area of plate convergence of plate tectonics as wide zones of deformation, and those and deformation, but the location of plate boundaries boundaries involving a continental plate boundary are much remains uncertain. At least two microplates, Panama and wider than are those consisting only of oceanic plates North Andes, have been hypothesized by Kellogg et al. 6Isaacs et al., 1968; Dewey and Bird, 1970; Freymueller et 61985) and Kellogg and Vega 61995) 6Figs. 1 and 2). This al., 1993; Kellogg and Vega, 1995; Gutscher et al., 1999). paper presents the results of CASA geodetic measurements Global positioning system 6GPS) measurements provide spanning the years 1991±1998 in Costa Rica, Panama, cost-effective and precise constraints on models of plate Colombia, Ecuador, and Venezuela and focuses on GPS tectonic processes at convergent boundaries, such as conti- measurements related to the following: nuum deformation versus microplate or block rotation 6Thatcher, 1995). 1. Oblique subduction at the Ecuador trench and the `escape'of the North Andes; 2. Earthquake strain accumulation at the Ecuador trench; * Corresponding author. Tel.: 11-803-777-4501; fax: 11-803-777-9133. 3. Island arc±continent collision, Panama±Colombia; and E-mail address: [email protected] 6R. Trenkamp). 4. Caribbean plate subduction.

Fig. 1. Tectonic map for the CASA GPS project area. Seismicity with magnitudes greater than 4 between 1973 and 1999 6NEIC/USGS) are plotted with small black asterisks. Focal mechanisms 6Harvard CMT) are plotted for major earthquakes with their magnitudes. The large earthquakes are referenced to Table 1 by the numbers superimposed on the compressional quadrant.

2. Data analysis ®nd and correct cycle-slips 6phase breaks in the data stream) and remove outliers. The data then are decimated to a 6- All data presented in this paper have been analyzed using minute interval for data collected before 1996 and a 5-minute GIPSY OASIS or GYPSY OASIS II software developed at interval for data collected after 1996 to enable the use of the the Jet Propulsion Laboratory 6JPL), California Institute of JPL's precise clocks and orbits. The GIPSY OASIS II solu- Technology 6Lichten and Border, 1987; Blewitt, 1989, tion suite of software was then run following a strategy simi- 1990). Details of the 1991 data analysis are given in Frey- lar to that described by He¯in et al. 61992) to obtain solutions mueller 61991) and Freymueller et al. 61993). Each day of based on all data available for a given day. All daily solutions, 1991 data was analyzed independently using GIPSY OASIS, except those for 1991, were derived with weak constraints on the data from all stations in the CASA region, and data from initial positions and then transformed into ITRF96 using tracking sites distributed over almost half the globe. The JPL-produced daily frame ®les. The 1991 data were trans- 1994, 1996, and 1998 data were analyzed using GIPSY formed to ITRF96 using a seven-parameter transformation. OASIS II. All days were analyzed independently using all the stations in the CASA region and a tracking network similar in the number of stations to the 1991 campaign but with a 3. Velocity ®eld descriptions more favorable regional geographic distribution for CASA due to the increased number of permanent sites, especially in CASA GPS measurements used in this study were made the southern hemisphere. All the raw data were passed during 1991, 1994, 1996, and 1998. Forty-four usable through automatic editors: Turboedit 6Blewitt, 1990) for vectors for tectonic interpretation 6Table 2, Fig. 3) were Rogue and Turborogue receivers and Phasedit 6Freymueller, determined and are distributed over the major tectonic unpublished algorithm) for all other receivers. These editors features of the study area. Two of our sites are located on R. Trenkamp et al. / Journal of South American Earth Sciences 15 62002) 157±171 159

Fig. 2. Fault and tectonic feature map of the project area. A±A0 and B±B0 are the transects that were used for the modeling in Figs. 5 and 6. the Nazca plate, the International GPS Service permanent and 68N, the velocity vectors of the stations PAST, CALI, station on the Galapagos Islands 6GALA) and a station on BUEN, MZAL, and BOGO have a distinctly greater north- Malpelo Island 6MALS). The consistent velocities and ern component to their vectors. This suggests that the obli- directions for these two vectors indicate an approximately que component of the Nazca plate subduction is being due east-oriented motion relative to a South America ®xed accommodated by transpressive NE motion along faults reference frame 6South America Euler pole and velocity in subparallel to the Andean margin. Above latitude 68N, ITRF96; Edmundo Norabuena, personal communication, stations RION, MONT, CART, BUCM, VDUP, and URIB 2001) 6Fig. 3). have a greater eastward component to their velocity vectors, East of the Colombia±Ecuador trench, the coastal which is signi®cant to at least 500 km from the trench and stations MANT, PAJA, MUIS, ESME, and TUMA have probably related to the ongoing Panama±Colombia colli- vectors with a signi®cant eastern component, though sion 6Kellogg and Vega, 1995). Stations VDUP and much smaller than the velocity vector measured for the BUCM in Colombia, which are approximately 500 km subducting Nazca plate, whereas motions at the more north- from the collision zone, again show a signi®cant northern erly coastal sites, BUEN and BHSL, have a much smaller component along with the eastern component. Stations eastern component. The motion of the coastal sites suggests ELBA and MERI in Venezuela, which are much farther that some of the Nazca plate velocity is transferred directly from the Panama±Colombia collision zone, have similar to the overriding continental plate south of latitude 28N. The vectors to those between 28N and 68N. Five station vectors velocity vectors of the more eastern stations in Ecuador and in Panama, DAVI, CHIT, FLAM, ALBR, and CHEP, and southern Colombia decrease eastward until measured velo- three vectors in Costa Rica, LIMO, BRAT, and MANZ, give cities relative to stable South America are statistically not the relative magnitude and direction of the ongoing distinguishable from zero at LAGO, CONE, and VILL, east Panama±South America collision and imply relatively of the Eastern Andes frontal fault zone. uniform translation of the Panama block. Two sites in North and east of station TUMA, between latitudes 28N western Costa Rica 6LIBE and ETCG) have distinctly 160 R. Trenkamp et al. / Journal of South American Earth Sciences 15 62002) 157±171

Fig. 3. Station velocity vectors relative to stable South America at 95% con®dence using the data from the 1991, 1994, 1996, and 1998 CASA campaigns. The Cocos Island vector 6COCO) was calculated relative to South America using a vector reported in Freymueller et al. 61993). Location of the COCO vector has been shifted to maintain the clarity of the ®gure. different motions from those in eastern Costa Rica and strike-slip motion 6Henneberg, 1983). Based on earthquake Panama. These vectors provide some control on the western focal mechanisms Pennington 61981) proposes that trans- extent of the Panama microplate. The sole vector for the pressive right-lateral slip is occurring on the westward- stable Caribbean plate, SANA, suggests continuing slow dipping faults of the East Andean frontal fault system subduction of the Caribbean plate under northern Colombia while the North Andes Block is moving NNE relative to and Panama. the South American plate. The DGM is a system of northeast-trending, right-lateral strike-slip faults and north-trending thrust faults. Slip rates, based on morphology changes, 4. Discussion along a splay of the Pallatanga fault in the central Ecuador- ian Andes are reported as 3±4.5 mm/a 6Winter et al., 1993). 4.1. Oblique subduction at the Ecuador trench and escape of Seismic activity of the Iliniza fault, which is subparallel to the North Andes and approximately 70 km to the north of the Pallatanga fault, is similar to that of the Pallatanga fault and probably The North Andes Block is delineated by the Bocono fault, has a similar slip rate 6Hugo Yepes, Escuela Politecnica East Andean fault system and the Dolores Guayaquil mega- Nacional, Ecuador, personal communication, 1991). Slip shear 6DGM) to the east, the South Caribbean deformed rates are estimated as 7 ^ 3 mm/a along the Rio±Chin- belt on the north, and the Colombia±Ecuador trench and gual±La So®a fault at the Ecuador±Colombia border, on Panama on the west 6Bowin, 1976; Pennington, 1981; the basis of offsets of dated pyroclastic ¯ows 6Ego et al., Kellogg et al., 1985; Adamek et al., 1988; Ego et al., 1996). Tibaldi and Leon 62000) estimate a right-lateral 1996; Gutscher et al., 1999) 6Figs. 1 and 2). Schubert displacement rate of 15±16 mm/a in northern Ecuador and 61982) describes the Bocono fault in Venezuela as a right- 13 mm/a in southern Colombia using morphological offsets lateral strike-slip fault that runs through the Merida Andes. across several faults that correspond to the East Andean Small-scale geodetic networks in the Merida Andes indicate Frontal fault system. Earthquake slip rate estimates, offset compression normal to the Bocono fault, in addition to the glacial moraines, and Quaternary fault morphology suggest R. Trenkamp et al. / Journal of South American Earth Sciences 15 62002) 157±171 161

Fig. 4. Observed vectors relative to South America in the area of the great 1906 earthquake, which ruptured an over 500 km-long segment of the Colombia± Ecuador forearc. The 1942, 1958, and 1979 earthquake sequence reruptured the 1906 rupture zone. Rupture zones shown are after Kanamori and McNally 61982). that the northern Andes are moving northeastward relative geomorphic observations 6Tibaldi and Leon, 2000). This to stable South America at 4±10 mm/a along the Bocono slowing may be related to the slowing Nazca±South fault in the Merida Andes of Venezuela 6Schubert, 1980, America subduction rates during the same period of time 1982; Aggarwal, 1983; Lavenu et al., 1995). 6Norabuena et al., 1999). CASA GPS measurements are consistent with rapid Transition of the boundary of the North Andes Block subduction of the Nazca plate and Carnegie aseismic ridge from the DGM in the south to the East Andean Frontal at the Ecuador trench beneath stable South America and Fault system takes place on faults in and to the north of show that the direction of this motion is oblique to the the Inter-Andean Valley 6Lavenu et al., 1995; Ego et al., Colombia±Ecuador margin. As a result, the Andean margin 1996; Gutscher et al., 1999). Four CASA sites 6JERS, is undergoing shortening perpendicular to the margin, and LATA, AMBT, and RIOP) are located from north to south as predicted, the North Andes Block appears to be `escap- within this transition zone. The measured velocity vectors ing' to the northeast parallel to the margin 6Jarrard, 1986; rotate clockwise from north to south 6Fig. 4). These motions Beck, 1991; McCaffrey, 1992). Relative to station BOGO, are interpreted to indicate distributed right-lateral motion seven stations in the North AndesÐLATA, JERS, PAST, along a fault zone between RIOB and JERS or elastic defor- BUEN, BUCM, URIB, and MERIÐhave statistically zero mation from a single locked fault. motion at two sigma. The velocity at BOGO relative to The large eastern component of motion at sites in this South America 66 ^ 2 mm/a) is therefore interpreted as transition zone is hypothesized to be due to permanent approximating the velocity of the North Andes Block shortening related to the East Andean Frontal Fault system. relative to stable South America. This rate is less than the The residual motions of other sites on the North Andes average North Andes±South America transcurrent motion Block, which indicate deviations from coherent rigid plate rate for the Pleistocene, as measured by sur®cial behavior, are discussed in detail later. 162 R. Trenkamp et al. / Journal of South American Earth Sciences 15 62002) 157±171

Table 1 Summary information for the major earthquakes shown in Figs. 1 and 4 6source 1: Harvard CMT solution 2000; source 2: Kanamori and McNally 61982))

20 Date Longitude 6degrees) Latitude 6degrees) Depth 6km) Mo 610 Nm) Ms Source

North Andes earthquakes 01/31/06 279.00 20.75 25.0 20,000.00 8.7 2 05/14/42 280.00 1.37 25.0 320.00 7.9 2 01/19/58 280.60 1.00 60.0 520.00 7.8 2 1 04/09/76 280.11 0.79 19.4 11.10 6.7 1 2 12/12/79 281.19 2.32 19.7 1,690.00 7.7 1 3 05/06/81 279.03 21.99 17.4 5.60 6.4 1 4 11/22/83 280.01 0.31 35.2 16.60 6.2 1 5 03/06/87 282.16 20.06 15.0 63.70 6.9 1 6 09/22/87 281.74 20.89 15.0 4.07 6.2 1 7 11/19/91 282.82 4.80 19.1 73.20 7.0 1 8 10/17/92 283.61 7.22 15.0 11.30 6.7 1 9 10/18/92 283.66 7.27 15.0 57.10 7.3 1 10 11/04/96 282.79 7.47 15.0 3.05 6.0 1 11 08/04/98 279.52 20.57 25.6 63.70 7.1 1 East Andean Frontal Fault earthquakes 12 01/19/95 287.17 5.16 16.0 7.07 6.6 1 13 10/03/95 282.47 22.55 25.0 39.10 7.0 1 14 10/03/95 282.23 22.88 15.0 5.38 6.1 1 Costa Rican earthquakes 15 03/25/90 275.42 9.95 17.9 110.00 7.0 1 16 04/22/91 277.23 10.10 15.0 331.00 7.5 1 17 02/09/92 272.19 11.20 15.0 340.00 7.2 1 18 08/20/99 275.90 9.28 24.0 26.00 6.9 1

4.2. Earthquake strain accumulation at the Ecuador trench ability for a large Ms . 7:5 earthquake in the same area during the 10-year period ending in 2002. Rapid subduction of the oceanic Nazca plate and aseismic Several sites of the CASA network are well situated to Carnegie ridge 6Fig. 3) 658 ^ 2 mm/a) is occurring at the measure the strain release in the event of a large earthquake Colombia±Ecuador trench. The forearc north of the Carne- in the area of the 1906 rupture zone and, more important, the gie ridge is historically seismically active 6Kelleher, 1972; strain that must be accumulating that would be capable of Kanamori and McNally, 1982; Mendoza and Dewey, 1984; producing such an event. Three CASA vectors at sites near Ruff, 1996; Swenson and Beck, 1996). During the 20th the coast at the Ecuador±Colombia border, at Muisne 6MUIS) century, a sequence of four great Ms . 7:5; subduction- and Esmeraldas 6ESME) in Ecuador and Tumaco 6TUMA) in related earthquakes occurred. The ®rst and largest event of Colombia, have large east±northeast motion relative to stable this series occurred in 1906 Mw 8:8 and, according to South America 6Figs. 3 and 4). Motion at the two sites in intensity reports and strong macroseismic activity, had a Ecuador average 21 mm/a relative to South America, whereas rupture length of 500 km 6Kelleher, 1972; Kanamori and motion at the more northerly site in Colombia is 15 mm/a McNally, 1982). Kanamori and McNally 61982), using relative to South America. Differences in the rates at these aftershock sequences to determine rupture areas, ®nd that three sites are probably purely related to distance from the three smaller events in 1942 Mw 7:9; 1958 Mw 7:8; trench, as all are within the rupture zone of the 1906 earth- and 1979 Mw 8:2 appear to have reruptured most of the quake. The direction of the movement of these sites relative to thrust fault plate boundary segment that ruptured during the stable South America is consistent with the direction of the 1906 event. However, the total of the seismic moments of subducting Nazca plate and Carnegie ridge, but the vectors the three events was barely 15% of the moment release have a smaller magnitude, which suggests that the motion during the 1906 event 6Table 1) 6Kanamori and McNally, re¯ects the partial transfer of motion of the subducting slab 1982; Mendoza and Dewey, 1984). Nishenko 61989), study- to the overriding continental plate. We hypothesize that the ing earthquake zones on the Paci®c Rim, estimated a 60± velocities of these coastal sites relative to South America 90% time-dependent probability for the recurrence of either re¯ect two modes of deformation approximately parallel to a large 7:0 , Ms , 7:7 or great Ms . 7:7 shallow-plate the direction of Nazca±South America convergence: boundary earthquake during the 10-year period 1989±1999, and Papadimitriou 61993), studying only segments along the 1. Deformation due to elastic recoverable strain accumulation western coast of South and Central America and using a associated with the seismic cycle at a locked or partially time-predictable model, gave a 68% time-dependent prob- locked subduction zone interface, and R. Trenkamp et al. / Journal of South American Earth Sciences 15 62002) 157±171 163

Fig. 5. a. CASA geodetic data at one standard error compared with elastic half-space models. A convergence rate of 58 mm/a was determined by repeat measurements of the Galapagos site 6GALA). Shortening is modeled as 6 mm/a distributed as 3 mm/a on an east-dipping Inter-Andean fault 6IAF) and 3 mm/a on a generalized west-dipping thrust 6OF) in the Oriente Basin. The IAF dips at about 208 to the east and is assumed locked from 0 to 10 km depth 6Lavenu et al., 1995). The modeled eastern thrust fault dips at 308 to the west and is locked from 0 to 20 km depth. Aseismic slip refers to that percentage of the slip accommodated aseismically at the subduction zone. b. Fault model for cross section B±B0. Line is located in Fig. 2.

2. Motion due to permanent shortening associated with mined by Kelleher 61972) and Kanamori and McNally active faulting in the Inter-Andean and East Andean 61982). Tichelaar and Ruff 61991) estimated similar lock- frontal fault systems several hundred kilometers east of ing depths 636±53 km) for subduction zones in the the subduction zone. central and southern Andes. Combining the approach of Savage 61983) with a bound- To test this hypothesis, the elastic strain in northern ary element modeling program 63D-DEF; Gomberg and Ecuador and southern Colombia due to a locked subduc- Ellis, 1994), eastward surface strain is modeled by imposing tion zone along the Colombia±Ecuador trench can be back slip on a velocity ®eld de®ned by the plate conver- estimated using an elastic half-space model for a locked gence, similar to the approach outlined by Lef¯er et al. dipping thrust fault 6Savage, 1983). The estimation of the 61997). The rate determined by the CASA project at dip related to the locked seismic zone 6Fig. 5b) is based GALA 658 mm/a) is used to simulate the locked segment. on seismicity 6Pennington, 1981) and gravity and seismic Various models with 40 and 50 km locking depths were run refraction models 6Meissner et al., 1976; Kellogg and to compare models in which the entire east±west conver- Vega, 1995). Maximum locking depths are estimated at gence is accommodated on the locked subduction interface 40 km for the 1942, 1958, and 1979 earthquakes and at 60% aseismic slip), models in which part of the slip on the 50 km for the 1906 earthquake, following those deter- subduction interface is accommodated aseismically 6partial 164 R. Trenkamp et al. / Journal of South American Earth Sciences 15 62002) 157±171

Table 2 CASA vectors relative to South America ®xed reference frame, showing east and north components with one sigma errors

Site Longitude Latitude East 6mm/a) North 6mm/a) s6E) 6mm/a) s6N) 6mm/a)

ALBR 280.44 8.99 22.06 20.57 1.49 1.02 AMBT 281.35 21.26 9.46 22.28 3.14 1.24 BARI 289.50 8.48 3.08 20.74 7.17 2.82 BHSL 282.61 6.20 12.67 3.62 3.85 1.64 BOGO 285.92 4.87 5.78 1.29 1.99 1.14 BRAT 277.11 9.55 24.99 8.69 3.43 1.62 BUCM 286.82 7.12 9.27 2.22 2.32 1.14 BUEN 283.01 3.82 4.51 4.49 3.20 1.42 CALI 283.64 3.50 0.08 5.13 2.29 1.20 CART 284.50 10.36 15.10 22.34 3.24 1.32 CHEP 281.96 8.25 29.04 2.83 4.94 1.79 CHIT 279.59 7.99 30.48 0.69 5.58 1.83 CONE 283.15 0.24 1.12 21.72 5.49 2.35 DAVI 277.56 8.41 31.80 7.33 4.12 1.61 ELBA 288.89 9.18 0.48 6.85 6.59 2.58 ESME 280.37 0.99 20.83 3.20 2.74 1.33 ETCG 275.89 10.00 17.49 2.41 2.97 1.44 FLAM 280.48 8.91 24.20 0.86 3.70 1.91 GALA 269.70 20.74 58.18 20.01 1.44 1.08 JERS 281.64 20.01 7.48 1.78 2.08 1.12 JUNQ 292.91 10.46 21.78 0.31 6.78 2.38 LAGO 283.14 0.10 0.59 21.20 5.61 2.62 LATA 281.37 20.81 6.51 20.03 4.29 1.88 LIBE 274.57 10.65 14.32 4.05 2.09 1.21 LIMO 276.97 9.96 26.43 21.66 3.08 1.70 MALS 278.39 4.00 53.62 1.43 2.09 1.21 MANT 279.33 20.94 20.58 1.88 7.93 3.99 MANZ 277.32 9.62 29.86 7.38 3.65 1.77 MERI 289.13 8.79 4.99 1.66 3.50 1.53 MONT 284.32 8.89 16.51 20.86 2.43 1.24 MUIS 279.98 0.60 23.65 1.06 3.64 1.83 MULM 281.46 21.44 10.84 26.09 5.09 2.43 MZAL 284.53 5.03 17.92 4.75 3.58 1.73 PAJA 279.57 21.55 16.98 20.23 5.67 2.62 PAST 282.74 1.22 4.68 3.51 3.20 1.41 PPYN 283.42 2.48 0.47 5.25 4.70 2.33 RION 284.57 6.18 14.09 2.38 6.11 2.76 RIOP 281.35 21.65 8.28 24.24 5.12 2.64 SANA 278.27 12.52 19.18 24.49 1.65 1.11 TUMA 281.25 1.81 16.08 20.42 3.76 1.91 URIB 288.26 7.91 4.66 21.58 5.20 2.24 VDUP 286.75 10.44 10.10 4.02 3.14 1.30 VILL 286.62 4.07 0.00 22.53 2.94 1.29 VUEL 276.15 9.62 26.64 9.38 5.02 1.88 decoupling), and models in which some of the convergence 60% aseismic slip) overestimate the GPS observations in the is accommodated on thrust faults within and east of the near trench stations. Models without permanent shortening Andes mountains 6Figs. 2 and 5) 6Lavenu et al., 1995). underestimate the GPS observations for the eastern stations. The CASA stations were projected parallel to the trench, A model with 50% locking 629 mm/a) plus 6 mm/a short- to the east±west pro®le 6B±B0, Fig. 2), so that the east±west ening on the eastern faults ®ts most of the observations well distances between trench and stations were conserved. at 1 sigma con®dence 6Fig. 5a). These results are similar to Shortening was modeled as 6 mm/a distributed as 3 mm/a those obtained by Norabuena et al. 61998) for the central on an east-dipping fault 6IAF) within the Inter-Andean Andes of Peru and Bolivia, who measured approximately depression, approximately 320 km from the trench, dipping 30±40 mm/a of locking and 10±15 mm/a of crustal short- at 188E, and locked from 0 to 10 km, and on a generalized ening at the sub-Andean foreland fold and thrust belt. At the west-dipping thrust 6OF) in the Oriente Basin dipping 308W northern end of the 1906 rupture zone 6Fig. 4), no signi®cant and locked from 0 to 20 km 6Fig. 5b). locking is required to explain the motions of Buenaventura The range of realistic locking depths seems to have a 6BUEN), Cali 6CALI), and Popayan 6PPYN), Colombia minimal effect on the model results. Fully locked models 6Fig. 5a). R. Trenkamp et al. / Journal of South American Earth Sciences 15 62002) 157±171 165

Fig. 6. a. CASA geodetic data at one standard error compared with elastic half-space models with various shallow dip angles on the AUFZ. The 25 mm/a Panama±South America convergence rate was determined using the average of four stations from the relatively aseismic stable Panama: CHIT, ALBR, FLAM, and CHEP. The displacement includes 3 mm/a shortening on a generalized thrust fault in the EAF system and 3 mm/a eastward component of the strike-slip motion associated with the escape of the North Andes Block. b. East±west cross section A±A0 of the collisional zone. Pro®le location is shown in Fig. 2.

Our shortening estimate 66 mm/a) is higher than the slow 4.3. Island arc±continent collision, Panama±Colombia rate of Andean shortening inferred from seismic data 61.4± 2.1 mm/a) by Suarez et al. 61983). The seismic estimate Panama and northern Colombia provide a unique oppor- does not include aseismic deformation, and the sampling tunity to study the kinematics and mechanics of an active period may have missed infrequent, very large earthquakes. island arc±continent collision zone. The timing of the initial The observed strain accumulation and time since the last collision of Panama with northern Colombia is not well great earthquake in the region support the high conditional constrained, and different models span an interval of time probability of a large or great earthquake in the near future from 12 to 20 Ma 6Lonsdale and Klitgord, 1978; Keigwin, 6Nishenko, 1989; Papadimitriou, 1993). Assuming a 50% 1978, 1982; Wadge and Burke, 1983; Keller et al., 1989; locked plate interface since the last large earthquake on Duque-Caro, 1990; Kellogg and Vega, 1995). The Panama± the southern rupture zone in 1942, seismic slip has accumu- Colombia border area is a tectonically complex region that lated at a rate of approximately 29 mm/a for a total of 1.7 m. lies between the North Panama Deformed Belt and the If the accumulated slip is released as a single event, the Colombia±Ecuador trench 6Fig. 2) in which more than 60 expected magnitude would be Mw 7:5 2 7:7; using earthquakes of magnitude 5.0 or larger occurred between empirical scaling relations 6Kanamori, 1983), which 1963 and 1981. Studies of seismicity in the border region would be similar to the 1942 or 1958 events but smaller and offshore suggest that east±west compression has than the 1979 event. produced an area of diffuse and complex faulting along 166 R. Trenkamp et al. / Journal of South American Earth Sciences 15 62002) 157±171 northwest-to-northeast-trending strike-slip and thrust faults Proposed models for crustal deformation within Panama 6Pennington, 1981; Adamek et al., 1988). include a ¯exural beam 6Silver et al., 1990), distributed left- On October 17 and 18, 1992, a deadly sequence of two lateral slip 6Mann and Corrigan, 1990), a diffuse plate large shallow crustal earthquakes 6Ms 6:6 and Ms 7:3) boundary accommodating deformation over a broad area occurred near Murindo in northwestern Colombia 6Table 2 6Jordan, 1975; Wadge and Burke, 1983), and the interaction and Fig. 1, #8±9). The `best' ®tting focal mechanism for the of rigid microplates 6Pennington, 1981; Adamek et al., smaller event was a northeast-trending thrust event, whereas 1988; Kellogg et al., 1989; Kellogg and Vega, 1995). the larger event was associated with strike-slip motion on an Relative to ALBR, vectors in Panama suggest that the inferred north-trending fault 6Li and Toksoz, 1993; Wallace central, virtually aseismic section of Panama 6Figs. 1 and and Beck, 1993; Ammon et al., 1994). These focal mechan- 3) is moving rigidly, whereas the DAVI vector in western isms are consistent with compression normal to the Panama shows a northeast velocity. The vector at DAVI can Panama±North Andes suture 6Freymueller et al., 1993). be explained as elastic strain accumulation from the locked Evidence for active Panama±North Andes collision was subduction of the Cocos Ridge under Central America ®rst reported by Kellogg and Vega 61995), and estimates of 6Lundgren et al., 1999). the total shortening exceed 150 km 6Colletta et al., 1990). Seismicity and GPS measurements in Panama and north- Relative to a ®xed South America reference frame, four ern Colombia neither support an actively ¯exing Panama well-determined vectors in Panama 6CHIT, ALBR, FLAM, 6Silver et al., 1990), nor an active left-lateral shearing and CHEP) demonstrate the ongoing collision of the Panama 6Mann and Corrigan, 1990), though geology suggests that microplate with the North Andes Block at an average of both have been important deformation mechanisms in the 25 mm/a. The large eastern components of the vectors in geologic past. Seismicity 6Marshall et al., 2000) and GPS northern Colombia 6CART, MONT, RION, VDUP, and results con®rm a rigid Panama microplate that includes BUCM) suggest the transfer of motion from the Panama much of Costa Rica, as has been proposed by several microplate collision. In the ongoing collision, the mechanical authors 6Kellogg and Vega, 1995; Lundgren et al., 1999) behavior of the Panama arc can be approximated as a rigid 6Fig. 1). The measured velocities of the sites on the Nazca indenter. The wide continental plate boundary zone in Colom- plate relative to the Panama microplate 6Fig. 3) are consis- bia appears to be compressing normal to the suture zone at the tent with left-lateral slip on a fault south of Panama, as Panama±Colombia boundary. Monteria 6MONT) 6Fig. 3), proposed by Jordan 61975) and described by Westbrook et 120 km east of the suture, is moving eastward at al. 61995). The relative velocities of the sites on the Nazca 16 ^ 3 mm/a. The deformation appears to extend up to plate 6MALS and GALA) are consistent with a single Euler 600 km from the suture into the continental crust. vector for the Nazca plate 6Trenkamp and Kellogg, in prep.) Deformation in northern Colombia due to the Panama and do not require nonrigid behavior of the northern Nazca collision can also be approximated by elastic half-space plate, as was suggested by Wolters 61986) and Adamek et al. models for a locked dipping thrust fault 6Savage, 1983; 61988), and Kellogg and Vega 61995). The hypothesis of a Gomberg and Ellis, 1994). An estimated maximum locking separate microplate between the Nazca plate and the depth of 40 km is based on microseismicity studies near the Panama±Costa Rica microplate de®ned by motion along Panama±Colombia suture zone that determined changes in Hey's 61977) fault north of Malpelo Island cannot be tested focal mechanisms at approximately 36 km, coincident with with the available data. the crust mantle boundary 6Hutchings et al., 1981). All The northeastward displacement vectors for the three models assume that the entire east±west convergence is sites in southeastern Costa Rica 6Fig. 3) are indistinguish- accommodated on the locked interface 6no aseismic slip) able from the vector for David 6DAVI) in western Panama. at the Atrato±Uraba Fault Zone 6AUFZ) 6A±A0, Figs. 2 These displacements can be interpreted as the result of the and 6b). Shortening is estimated as 3 mm/a locking on thrust eastward motion of the Panama microplate, plus elastic faults in the East Andean Frontal fault 6EAF) system and locking associated with the subduction of the Cocos 3 mm/a through-going displacement of the strike-slip Ridge. The anomalous Limon 6LIMO) displacement may motion associated with the northward-directed escape of include post-seismic slip from the 1991 earthquake. The the North Andes Block. two sites in northwest Costa Rica, LIBE and ETCG, are Models without shortening do not ®t the observed east- moving in a distinct WNW direction relative to the Panama ward displacements of BUCM, VDUP, and URIB 6Fig. 6a). microplate. We therefore locate the western boundary of the To model the motion observed at URIB, permanent short- Panama microplate between ETCG and VUEL. This bound- ening and northeastward escape on eastern faults are esti- ary coincides with a seismic zone of left-lateral strike-slip mated to be about 6 mm/a. Near ®eld sites MONT, CART, motion recognized by Fan et al. 61993). and RION are well modeled by a locked low-angle 613±158 thrust fault 6AUFZ). The Valedupar 6VDUP) and Bucara- 4.4. Caribbean plate subduction manga 6BUCM) vectors are not well matched by the models but may re¯ect unmodeled deformation produced by the The Caribbean plate interacts directly with at least three Panama collision and Caribbean oblique subduction. lithospheric plates: North America, South America, and R. Trenkamp et al. / Journal of South American Earth Sciences 15 62002) 157±171 167 Cocos 6Fig. 1), and many models have been proposed for bean subduction and Panama±North Andes collision or Caribbean neotectonics 6Jordan, 1975; Ladd, 1976; Burke et permanent eastward shear deformation of the South Carib- al., 1984; Stein et al., 1988; Mann et al., 1990; DeMets et al., bean deformed belt.. 1994; Deng and Sykes, 1995). Of these interactions, one of CASA GPS measurements show a 7 ^ 2 mm/a south- the most controversial problems in Caribbean tectonics is westward convergence between the Caribbean plate the location of the Caribbean±South American plate bound- 6SANA) and the Panama microplate. This southwest ary. Several studies of historical seismicity place the south- convergence produced the April 1991 Costa Rica earth- western Caribbean±South American boundary along the quake Ms 7:5 6Goes et al., 1993; Lundgren et al., Bocono fault in the Venezuelan Andes 6Dewey, 1972; 1993; Protti and Schwartz, 1994; Suarez et al., 1995), active Aggarwal, 1983). Focal mechanisms along the southeastern folding in the north Panama deformed belt 6Silver et al., Caribbean margin suggest right-lateral transpression with 1995), and a south-dipping Wadati±Benioff zone beneath slow abduction of the southern Caribbean plate onto South Panama 6Adamek et al., 1988; Protti et al., 1994; Mann and America 6Perez and Aggarwal, 1981; Speed, 1985). Seismic Kolarsky, 1995; Silver et al., 1995). re¯ection and gravity surveys expose a major folded sedi- Pleistocene activity on the Santa Marta±Bucaramanga mentary basin south of the Venezuelan and Colombian fault system 6Fig. 2) is shown by deformed terraces and topographic basins from the Magdalena Delta to the Los offset stream patterns 6Campbell, 1968). Although we do Roques Canyon 6Edgar et al., 1971; Case, 1974). Multi- not have direct measurements across the Santa Marta± channel seismic re¯ection pro®les across this south Carib- Bucaramanga fault, a relative left-lateral displacement of bean deformed belt show that Caribbean acoustic basement up to 6 ^ 2 mm/a is possible. This estimate is the vector has underthrust the deformed belt 6 Silver et al., 1975; difference between Cartagena 6CART) and Valledupar Bowin, 1976; Talwani et al., 1977; Lu and McMillen, 6VDUP) and between Bucaramanga 6BUCM) and El 1982; Lehner et al., 1983; Ladd et al., 1984). Folding of Batey 6ELBA) parallel to the fault. the youngest sediments and the presence of a large isostatic negative gravity anomaly over the Curacao Ridge suggest active deformation. Furthermore, Dewey 61972) and 5. Summary and conclusions Pennington 61981) recognized a zone of earthquakes dipping about 208 to the southeast and terminating 200 km CASA GPS measurements show that rapid subduction below the Maracaibo Basin. Kellogg and Bonini 61982) and of the Nazca plate is occurring along the length of the Toto and Kellogg 61992) interpreted these earthquakes as a Colombia±Ecuador trench. The subduction appears to be Benioff zone produced by slow amagmatic subduction of accommodated differently from south to north along the Caribbean lithosphere. This interpretation is supported by trench, probably due to the in¯uence of the subduction of seismic tomographic evidence for a south-dipping, high the Carnegie aseismic ridge in Ecuador and southern velocity slab 6van der Hilst and Mann, 1994), studies of Colombia. Coastal sites in Ecuador and southern Colom- intermediate depth seismicity 6Malave and Suarez, 1995), bia 6MANT, PAJA, MUIS, ESME, and TUMA) show microseismicity 6Perez et al., 1997), and the structure of signi®cant transfer of the motion of the subducting slab northern and eastern Venezuela 6Lallemant, 1997). to the overriding plate and are characterized by large CASA GPS measurements show oblique east±southeast eastward components to their relative motion vectors. convergence of 20 ^ 2 mm/a between the Caribbean island These vectors are hypothesized to re¯ect two modes of San Andres 6SANA) and stable South America 6Fig. 3). This deformation, elastic recoverable strain and active faulting result is consistent with the angular velocity vector for associated with permanent shortening, and suggest a Caribbean±South American relative motion based on GPS possible rationale for the repeat sequences of great earth- results from eight Caribbean sites and ®ve South American quakes. Sites in central western Colombia 6BUEN and sites 651.58N, 265.78E, 0.2728/m.y.) of Weber et al. 62001). BHSL) have a minimal eastern component to their Our results and the Weber et al. 62001) velocity vector are vectors and have not historically experienced great earth- also consistent with the GPS results of Perez et al. 62001) for quakes. the Caribbean±Venezuelan margin, which indicates east± Convergence rates along the Colombia±Ecuador trench west dextral displacement of about 20 mm/a. Lateral escape and the transferred motion to the North Andes, escaping of the North Andes Block to the northeast along the rigidly at 6 ^ 2 mm/a to the northeast, support the inference Bocono±East Andes±DGM fault system at 6 ^ 2 mm/a that present plate velocities have slowed relative to time- increases the southward component of the Caribbean± averaged plate motions 6Norabuena et al., 1999). The North Andes relative motion. The relative Caribbean observed permanent Andean shortening normal to the 6SANA)±North Andes 6BOGO) vector is 14 ^ 2 mm/a to margin and consistent northeastward escape of the northern the southeast. Southeastward motion at two sites on the Andes parallel to the margin for more than 1400 km from North Colombia accretionary prism, Monteria 6MONT) Ecuador to Venezuela suggests that either the margin paral- and Cartegena 6CART), relative to the North Andes lel strength of the North Andes is much greater than its 6BOGO) suggests elastic deformation produced by Carib- margin normal strength or, more likely, that the South 168 R. Trenkamp et al. / Journal of South American Earth Sciences 15 62002) 157±171 American craton is acting as a rigid buttress. The GPS Acknowledgements measurements to date are insuf®cient to demonstrate distinct North Andes and Maracaibo Blocks, as proposed by Mann Funding for this project was provided through National and Burke 61984). The measurements are consistent, Science Foundation grants EAR-8617485 and EAR- however, with possible left-lateral slip on the Santa 8904657 and NASA. The observations described herein Marta±Bucaramanga fault. were the cooperative efforts of many institutions. We espe- In northwestern Colombia and Panama, CASA measure- cially recognize the efforts of Dr Adolfo Alarco, INGEO- ments demonstrate the active collision of Panama with MINAS, and Instituto Geogra®co Augustin Codazzi in Colombia and ongoing deformation up to 550 km from Colombia; Colonel Anibal Salazar and Susana Arciniegas the suture zone. Velocity vectors within Panama suggest and the Instituto Geogra®co Militar in Ecuador; Instituto that the Panama microplate is behaving as an approximately Geogra®co Tommy Guardia in Panama; Paul Lundgren, rigid indenter. Combined with the work of Lundgren et al. JPL, for the Costa Rica data; and Hermann Drewes, 61999) and Marshall et al. 62000), CASA GPS results Deutsches Geodtisches Forschungsinstitute, for the Vene- con®rm a rigid Panama microplate whose western boundary zuela data. We thank Tim Dixon, Paul Mann, Hugh is in central Costa Rica. The velocities of the two vectors on Cowan, and Susan Beck for careful reviews that improved the Nazca plate 6GALA and MALS) relative to the Panama the paper. Many other persons and institutions who microplate are consistent with left-lateral slip on a fault contributed to the CASA GPS project are listed in Kellogg south of Panama and do not require a separate north et al. 61989). Figures 1±4 created with Generic Mapping Nazca microplate. The Caribbean±South American conver- Tool 6GMT) 6Wessel and Smith, 1995). gence rate is 20 ^ 2 mm/a. Various driving mechanisms have been proposed for the northeastward escape of the North Andes. Pindell References and Dewey 61982) proposed that the Panama collision is driving North Andean/Maracaibo escape. This mechan- Adamek, S., Frohlich, C., Pennington, W.D., 1988. Seismicity of the Carib- ism is unlikely to be signi®cant in Ecuador and southern bean±Nazca boundary; constraints on microplate tectonics of the Colombia, however. Russo and Silver 61995), based on Panama region. J. Geophys. Res. 93, 2053±2075. evidence of seismic shear wave splitting, suggested that Aggarwal, Y., 1983. Seismic slip rates and earthquake rupture zones in the mantle ¯ow from the Paci®c into the Caribbean is inter- southern Caribbean: implications for plate motions and earthquake acting with subducting mantle slabs. According to hazard in this region. Abstracts of the 10th Caribbean Geological Conference, p. 16. Pennington 61981) and Gutscher et al. 61999), the arrival Ammon, C.J., Lay, T., Velasco, A.A., Vidale, J.E., 1994. Routine estima- of the Carnegie Ridge at the Ecuador trench 6Fig. 1) tion of earthquake source complexity; the 18 October 1992 Colombian initiated the escape of the North Andes. Kellogg and earthquake. Bull. Seismol. Soc. Am. 84, 1266±1271. Mohriak 62001) proposed that the rapid oblique subduc- Beck Jr, M.E., 1991. Coastwise transport reconsidered, lateral displace- tion of the Nazca plate, as well as the Carnegie Ridge, ments in oblique subduction zones and tectonic consequences. Phys. Earth Planet. Int. 68, 1±8. may drive the North Andean detachment. McCaffrey Blewitt, G., 2000. Carrier phase ambiguity resolution for the Global Posi- 61994) noted that greater plate obliquities 6Sumatra, tioning System applied to geodetic baselines up to 2000 km. J. New Zealand, Aleutians) generally correlate with greater Geophys. Res. 94, 10,187±10,203. arc-parallel slip partitioning. Geodynamic models are Blewitt, G., 1990. An automatic editing algorithm for GPS data. Geophys. needed to determine whether mantle ¯ow is a potential Res. Lett. 17, 199±202. Bowin, C., 1976. Caribbean gravity ®eld and plate tectonics. Spec. Pap.Ð driving mechanism. Ridge subduction may also be Geol. Soc. Am. 169, 79. important. The uncoupling of Panama from the Carib- Burke, K., Cooper, C., Dewey, J.F., Mann, P., Pindell, J.L., 1984. Carib- bean plate and its eastward collision with the North bean tectonics and relative plate motions. The Caribbean±South Amer- Andes may have been driven by oblique subduction of ican plate boundary and regional tectonics, Bonini, W.E., Hargraves, the Cocos Ridge and the Nazca and Cocos plates 6Mann R.B., Shagam, R. 6Eds.). Mem.ÐGeological Soc. Am. 162, 31±63. Campbell, C.J., 1968. The Santa Marta wrench fault of Colombia and its and Kolarsky, 1995). regional setting: transactions. Caribbean Geological Conference IV, Continued GPS measurements are needed to differentiate Trinidad, 1965, pp. 247±260. between permanent deformation and transient visco-elastic Case, J.E., 1974. Major basins along the continental margin of northern deformation of the Andean margin. Site densi®cation will South America. In: Burk, C.A., Drake, C.L. 6Eds.). The Geology of show whether permanent deformation is con®ned to rigid Continental Margins. pp. 733±741. Chase, C.G., 1978. Plate kinematics: the Americas, East Africa, and the rest blocks 6abrupt changes in direction and velocity of vectors) of the world. Earth Planet. Sci. Lett. 37, 355±368. or is a smoothly varying continuum. Particularly intriguing Colletta, B., Gebrard, F., Letouzey, J., Werner, P., Rudkiewitz, J.L., 1990. is the crustal deformation associated with the subduction of Tectonic style and crustal structure of the Eastern Cordillera 6Colom- the Carnegie Ridge. For example, preliminary GPS bia) from a balanced cross-section. Petroleum and Tectonics in Mobile measurements suggest that seismic locking, and therefore Belts, Letouzey, J.P. 6Ed.). Editions Technip, 81±100. DeMets, C., Gordon, R.G., Argus, D.F., Stein, S., 1990. Current plate seismic hazard, is present in southern Ecuador, an area with motions. Geophys. J. Int. 101, 425±478. no great historic earthquakes. DeMets, C., Gordon, R.G., Argus, D.F., Stein, S., 1994. Effect of recent R. Trenkamp et al. / Journal of South American Earth Sciences 15 62002) 157±171 169

revisions to the geomagnetic reversal time scale on estimates of current Kelleher, J.A., 1972. Rupture zones of large South American earthquakes plate motions. Geophys. Res. Lett. 21, 2191±2194. and some predictions. J. Geophys. Res. 77, 2087±2103. Deng, J., Sykes, L.R., 1995. Determination of Euler Pole for contemporary Keller, G., Zenker, C.E., Stone, S.M., 1989. Late Neogene history of the relative motion of Caribbean and North American plates using slip Paci®c±Caribbean gateway. J. South Am. Earth Sci. 2, 73±108. vectors of interplate earthquakes. Tectonics 14, 39±53. Kellogg, J.N., Bonini, W.E., 1982. Subduction of the Caribbean Plate and Dewey, J.W., 1972. Seismicity and tectonics of western Venezuela. Bull. basement uplifts in the overriding South American Plate. Tectonics 1, Seismol. Soc. Am. 62, 1711±1751. 251±276. Dewey, J.F., Bird, J.M., 1970. Mountain belts and the new global tectonics. Kellogg, J., Mohriak, W., 2001. Tectonic and geological environment of J. Geophys. Res. 75, 2625±2647. coastal South America. In: Seeliger, U., Kjerfve, B. 6Eds.). Ecological Duque-Caro, H., 1990. The Choco Block in the northwestern corner of Studies. Coastal Marine Ecosystems of Latin America, vol. 144. South America; structural, tectonostratigraphic, and paleogeographic Springer, Berlin, pp. 1±16. implications. J. South Am. Earth Sci. 3, 71±84. Kellogg, J.N., Vega, V., 1995. Tectonic development of Panama, Costa Edgar, N.T., Ewing, J.I., Hennion, J., 1971. Seismic refraction and re¯ec- Rica, and the Colombian Andes: constraints from global positioning tion in Caribbean sea. Am. Assoc. Petrol. Geol. Bull. 55, 833±870. system geodetic studies and gravity. In: Mann, P. 6Ed.). Geologic and Ego, F., Sebrier, M., Lavenu, A., Yepes, H., Egues, A., 1996. Quaternary Tectonic Development of the Caribbean plate Boundary in Southern state of stress in the Northern Andes and the restraining bend model for Central America. GSA Special Paper, 295. pp. 75±90. the Ecuadorian Andes. Geodynamics of the Andes, Dewey, J.F., Lamb, Kellogg, J.N., Ogujiofor, I.J., Kansaka, D.R., 1985. Cenozoic tectonics of S.H. 6Eds.). Tectonophysics 259, 101±116. the Panama and North Andes blocks. MemoriasÐCongreso Latinoa- Fan, G., Beck, S.L., Wallace, T.C., 1993. The seismic source parameters of mericano de Geologia 6, 34±49. the 1991 Costa Rica aftershock sequence: evidence for a transcurrent Kellogg, J.N., Dixon, T.H., Neilan, R.E., 1989. CASA; Central and South plate boundary. J. Geophys. Res. 98, 15759±15778. America GPS geodesy. Eos, Trans. AGU 70, 649±651 655±656. Freymueller, J.T., 1991. CASAÐCentral and South America GPS Ladd, J.W., 1976. Relative motion of South America with respect to North geodesy: Crustal motions determined from 1988 and 1990 epoch America and Caribbean tectonics. Geol. Soc. Am. Bull. 87, 969±976. measurements in Colombia, Costa Rica and Ecuador. PhD Dissertation. Ladd, J.W., Truchan, M., Talwani, M., Stoffa, P.L., Buhl, P., Houtz, R., University of South Carolina, Columbia, pp. 172. Mauffret, A., Westbrook, G.K., 1984. Seismic re¯ection pro®les across Freymueller, J.T., Kellogg, J.N., Vega, V., 1993. Plate motions in the North the southern margin of the Caribbean. The Caribbean±South American Andean Region. J. Geophys. Res. 98, 21853±21863. plate boundary and regional tectonics, Bonini, W.E., Hargraves, R.B., Goes, S.D.B., Velasco, A.A., Schwartz, S.Y., Lay, T., 1993. The April 22, Shagam, R. 6Eds.). Mem.ÐGeol. Soc. Am., 153±159.

1991, Valle de la Estrella, Costa Rica Mw 7:7 earthquake and its Lallemant, H.G.A., 1997. Transpression, displacement partitioning, and tectonic implications; a broadband seismic study. J. Geophys. Res. 98, exhumation in the eastern Caribbean/South American plate boundary 8127±8142. zone. Tectonics 16, 272±289. Gomberg, J., Ellis, M., 1994. Topography and tectonics of the central New Lavenu, A., Winter, T., Davila, F., 1995. A Pliocene±Quaternary compres- Madrid seismic zone. J. Geophys. Res. 99 620), 20,299±20,310. sional basin in the Interandean Depression, central Ecuador. Geophys. Gutscher, M.A., Malavieille, J., Lallemand, S., Collot, J.Y., 1999. Tectonic J. Int. 121, 279±300. segmentation of the North Andean margin; impact of the Carnegie Lef¯er, L., Stein, S., Mao, A., Dixon, T., Ellis, M.A., Ocala, L., Selwyn, Ridge collision. Earth Planet. Sci. Lett. 168, 255±270. S.J., 1997. Constraints on present-day shortening rate across the Central He¯in, M., Beriger, W., Blewitt, G., Freedman, A.P., Hurst, K., Lichten, S., eastern Andes from GPS data, Geophys. Res. Lett. 24, 1031±1034. Lindqwister, U.J., Vigue, Y., Webb, F., Yunck, T.P., Zumberge, J., Lehner, P., Doust, H., Bakker, G., Allenbach, P., Gueneau, J., 1983. Active 1992. Global geodesy using GPS without ®ducial sites. Geophys. margins; Caribbean margin of South America, pro®les C-1422, C-1412, Res. Lett. 19, 131±134. and C-1413. In: Bally, A.W. 6Ed.). Seismic Expression of Structural Henneberg, H.G., 1983. Geodetic control of neotectonics in Venezuela. Styles; a Picture and Work Atlas. AAPG Studies in Geology, vol. 3. p. Tectonophysics 97, 1±15. 15 see also pp. 3.4.2-111±3.4.2-128. Hey, R., 1977. Tectonic evolution of the Cocos±Nazca spreading center. Le Pichon, X., 1968. Sea-¯oor spreading and continental drift. J. Geophys. Geol. Soc. Am. Bull. 88, 1404±1420. Res. 73, 3661±3697. van der Hilst, R., Mann, P., 1994. Tectonic implications of tomographic Li, Y., Toksoz, M.N., 1993. Study of the source process of the 1992 Colom-

images of subducted lithosphere beneath northwestern South America. bia Ms 7:3 earthquake with the empirical Green's function method. Geology 6Boulder) 22, 451±454. Geophys. Res. Lett. 20, 1087±1090. Hutchings, L., Turcotte, T., McBride, J., 1981. Microseismicity along the Lichten, S.M., Border, J.S., 1987. Strategies for high-precision global posi- Dolores shear zone in Antioquia, Colombia. Earthquake notes 52, 68± tioning system orbit determination. J. Geophys. Res. 92, 12,751± 69. 12,762. Isaacs, B., Oliver, J., Sykes, L.R., 1968. Seismology and the new global Lonsdale, P., Klitgord, K.D., 1978. Structure and tectonic history of the tectonics. J. Geophys. Res. 73, 5855±5899. eastern Panama Basin. Geol. Soc. Am. Bull. 89, 981±999. Jarrard, R.D., 1986. Terrane motion by strike-slip faulting of forearc slivers. Lu, R.S., McMillen, K.J., 1982. Multichannel seismic survey of the Colom- Geology 6Boulder) 14, 780±783. bia Basin and adjacent margins. Studies in Continental Margin Geol- Jordan, T., 1975. The present day motions of the Caribbean plate. J. ogy, Watkins, J.S., Drake, C.L. 6Eds.). AAPG Mem. 34, 395±410. Geophys. Res. 80, 4433±4439. Lundgren, P.R., Wolf, S.K., Protti, M., Hurst, K.J., 1993. GPS measure- Kanamori, H., 1983. Magnitude scale and quanti®cation of earthquakes. ments of crustal deformation associated with the 22 April 1991, Valle Tectonophysics 93, 185±199. de la Estrella, Costa Rica earthquake. Geophys. Res. Lett. 20, 407±410. Kanamori, H., McNally, K.C., 1982. Variable rupture mode of the subduc- Lundgren, P., Protti, M., Donnellan, A., He¯in, M., Hernandez, E., Jeffer- tion zone along the Ecuador±Colombia coast. Bull. Seismol. Soc. Am. son, D., 1999. Seismic cycle and plate margin deformation in Costa 72, 1241±1253. Rica; GPS observations from 1994 to 1997. J. Geophys. Res. 104 628), Keigwin, L.D., 1978. Pliocene closing of the Isthmus of Panama, based on 28,915±28,926. biostratigraphic evidence from nearby Paci®c Ocean and Caribbean Sea Malave, G., Suarez, G., 1995. Intermediate-depth seismicity in northern cores. Geology 6Boulder) 6, 630±634. Colombia and western Venezuela and its relationship to Caribbean Keigwin, L., 1982. Isotopic paleoceanography of the Caribbean and East Plate subduction. Tectonics 14, 617±628. Paci®c; role of Panama Uplift in late Neogene time. Science 217, 350± Mann, P., Burke, K., 1984. Neotectonics of the Caribbean. Rev. Geophys. 353. Space Phys. 22, 309±362. 170 R. Trenkamp et al. / Journal of South American Earth Sciences 15 62002) 157±171

Mann, P., Corrigan, J., 1990. Model for late Neogene deformation in tion and recurrence times. In: Bebout, G.E., Scholl, D.W., Kirby, S.H., Panama. Geology 6Boulder) 18, 558±562. Platt, J.P. 6Eds.). Subduction Top to Bottom. Geophysical Monograph, Mann, P., Kolarsky, R.A., 1995. East Panama deformed belt; structure, age, 96. pp. 91±104. and neotectonic signi®cance. Geologic and tectonic development of the Russo, R., Silver, P., 1995. The Andes' deep origins. Natural History 2, 53± Caribbean Plate boundary in southern Central America, Mann, P. 6Ed.). 58. Spec. Pap.ÐGeol. Soc. Am. 295, 111±130. Savage, J.C., 1983. A dislocation model of strain accumulation and release Mann, P., Schubert, C., Burke, K., 1990. Review of Caribbean neotectonics. at a subduction zone. J. Geophys. Res. 88, 4984±4996. In: Dengo, G., Case, J.E. 6Eds.). The Caribbean region, The Geology of Schubert, C., 1980. Late-Cenozoic pull-apart basins, Bocono fault zone, North America. pp. 307±338. Venezuelan Andes. J. Struct. Geol. 2, 463±468. Marshall, J.S., Fisher, D.M., Gardner, T.W., 2000. Central Costa Rica Schubert, C., 1982. Neotectonics of Bocono Fault, western Venezuela. deformed belt: kinematics of diffuse faulting across the western Panama Tectonophysics 85, 205±220. block. Tectonics 19, 468±492. Silver, E.A., Case, J.E., MacGillavry, H.J., 1975. Geophysical study of the McCaffrey, R., 1992. Oblique plate convergence, slip vectors and forearc Venezuelan borderland. Geol. Soc. Am. Bull. 86, 213±226. deformation. J. Geophys. Res. 97, 8905±8915. Silver, E.A., Reed, D.L., Tagudin, J.E., Heil, D.J., 1990. Implications of the McCaffrey, R., 1994. Global variability in subduction thrust zone±forearc north and south Panama thrust belts for the origin of the Panama systems. Pure Appl. Geophys. 142, 173±224. Orocline. Tectonics 9, 261±281. Meissner, R.O., Flueh, E.R., Stibane, F., Berg, E., 1976. Dynamics of the Silver, E.A., Galewsky, J., McIntosh, K.D., 1995. Variation in structure, active plate boundary in Southwest Colombia according to recent style, and driving mechanism of adjoining segments of the North geophysical measurements. Tectonophysics 35, 115±136. Panama deformed belt. Geologic and tectonic development of the Mendoza, C., Dewey, J.W., 1984. Seismicity associated with the great Caribbean plate boundary in southern Central America, Mann, P. Colombia±Ecuador earthquakes of 1942, 1958, and 1979; implications 6Ed.). Spec. Pap.ÐGeol. Soc. Am. 295, 225±233. for barrier models of earthquake rupture. Bull. Seismol. Soc. Am. 74, Speed, R.C., 1985. Cenozoic collision of the Lesser Antilles arc and conti- 577±593. nental South America and the origin of the El Pilar Fault. Tectonics 4, Minster, J.B., Jordan, T.H., 1978. Present day plate motions. J. Geophys. 41±69. Res. 83, 5331±5354. Stein, S., DeMets, C., Gordon, R.G., Brodholt, J., Argus, D., Engelin, J.F., Minster, J.B., Jordan, T.H., Molnar, P., Haines, E., 1974. Numerical model- Lundgren, P., Stein, C., Wiens, D.A., Woods, D.F., 1988. A test of ing of instantaneous plate tectonics. Geophys. J. R. Astr. Soc. 36, 541± alternative Caribbean Plate relative motion models. J. Geophys. Res. 576. 93, 3041±3050. Nishenko, S.P., 1989. Circum-Paci®c seismic potential 1989±1999. Open- Suarez, G., Molnar, P., Burch®el, C.B., 1983. Seismicity, fault plane File ReportÐUS Geological Survey, OF 89-0086, p. 126. solutions, depth of faulting, and active tectonics of the Andes of Norabuena, E., Lef¯er-Grif®n, L., Mao, A., Dixon, T., Stein, S., Sacks, I., Peru, Ecuador, and southern Colombia. J. Geophys. Res. 88, Selwyn, O.L., Ellis, M., 1998. Space geodetic observations of Nazca- 10,403±10,428. South America convergence across the central Andes. Science 279, Suarez, G., Pardo, M., Dominguez, J., Ponc, L., Montero, W., Boschini, I., 358±362. Rojas, W., 1995. The Limon, Costa Rica earthquake of April 22, 1991; Norabuena, E.O., Dixon, T.H., Stein, S., Harrison, C.G.A., 1999. Deceler- back arc thrusting and collisional tectonics in a subduction environ- ating Nazca±South America and Nazca±Paci®c plate motions. ment. Tectonics 14, 518±530. Geophys. Res. Lett. 26, 3405±3408. Swenson, J.L., Beck, S.L., 1996. Historical 1942 Ecuador and 1942 Peru Papadimitriou, E.E., 1993. Long-term earthquake prediction along the subduction earthquakes, and earthquake cycles along Colombia±Ecua- western coast of South and Central America based on a time predictable dor and Peru subduction segments. PAGEOPH 146, 67±101. model. PAGEOPH 140, 301±316. Talwani, M., Windisch, C.C., Stoffa, P.L., Buhl, P., Houtz, R.E., 1977. Pennington, W.D., 1981. Subduction of the eastern Panama Basin and Multichannel seismic study in the Venezuelan Basin and the Curacao seismotectonics of northwestern South America. J. Geophys. Res. 86, Ridge. In: Talwani, M., Pitman III, W.C. 6Eds.). Island Arcs, Deep Sea 10753±10770. Trenches and Back-Arc Basins. Maurice Ewing Series, vol. 1. pp. 83± Perez, O.J., Aggarwal, Y.P., 1981. Present-day tectonics of the southeastern 98. Caribbean and northeastern Venezuela. J. Geophys. Res. 86, 10791± Thatcher, W., 1995. Microplate versus continuum descriptions of active 10804. tectonic deformation. J. Geophys. Res. 100, 3885±3894. Perez, O.J., Jaimes, M.A., Garciacaro, E., 1997. Microseismicity Tibaldi, A., Leon, J.R., 2000. Morphometry of late Pleistocene±Holocene evidence for subduction of the Caribbean Plate beneath the faulting and volcanotectonic relationship in the southern Andes of South American Plate in northwestern Venezuela. J. Geophys. Colombia. Tectonics 19, 358±377. Res. 102 617), 17,875±17,882. Tichelaar, B.W., Ruff, L.J., 1991. Seismic coupling along the Chilean Perez, O.J., Bilham, R., Bendick, R., Velandia, J.R., HernaÂndez, N., subduction zone. J. Geophys. Res. 96, 11,997±12,022. Moncayo, C., Hoyer, M., Kozuch, M., 2001. Velocity ®eld across the Toto, El.A., Kellogg, J.N., 1992. Structure of the Sinu-San Jacinto fold belt; southern Caribbean plate boundary and estimates of Caribbean/South an active accretionary prism in northern Colombia. J. South Am. Earth American plate motion using GPS geodesy 1994±2000. Geophys. Res. Sci. 5, 211±222. Lett. 28, 2987±2990. Wadge, G., Burke, K., 1983. Neogene Caribbean plate rotation and asso- Pindell, J.L., Dewey, J.F., 1982. Permo±Triassic reconstruction of western ciated Central American tectonic evolution. Tectonics 2, 633±643. Pangea and the evolution of the Gulf of Mexico/Caribbean region. Wallace, T.C., Beck, S.L., 1993. The Oct. 17±18, 1992 Colombian earth- Tectonics 1, 179±211. quakes; slip partitioning or faulting complexity. Seismol. Res. Lett. 64, Protti, M., Schwartz, S.Y., 1994. Mechanics of back arc deformation in 29. Costa Rica; evidence from an aftershock study of the April 22, 1991, Weber, J.C., Dixon, T.H., DeMets, C., Ambeh, W.B., Jansma, P., Mattioli,

Valle de la Estrella, Costa Rica, earthquake Mw 7:7. Tectonics 13, G., Saleh, J., Sella, G., Bilham, R., PeÂrez, O., 2001. GPS estimate of 1093±1107. relative motion between the Caribbean and South American plates, and Protti, M., Guendel, F., McNally, K., 1994. The geometry of the Wadati± geological implications for Trinidad and Venezuela. Geology 29, 75± Benioff zone under southern Central America and its tectonic signi®- 78. cance; results from a high-resolution local seismographic network. Wessel, P., Smith, W.H.T., 1995. New version of the generic mapping tools Phys. Earth Planet. Int. 84, 271±287. released. EOS Trans. Am. Geophys. U. 76, 329. Ruff, L.J., 1996. Large earthquakes in subduction zones; segment interac- Westbrook, G.K., Hardy, N.C., Heath, R.P., 1995. Structure and tectonics R. Trenkamp et al. / Journal of South American Earth Sciences 15 62002) 157±171 171

of the Panama±Nazca plate boundary. In: Mann, P. 6Ed.). Geologic and measurements of displaced morphological features. Geophys. J. Int. Tectonic Development of the Caribbean Plate Boundary in Southern 115 63), 905±920. Central America. GSA Special Paper, 295. pp. 91±109. Wolters, B., 1986. Seismicity and tectonics of southern Central America Winter, T., Avouac, J.-P., Lavenu, A., 1993. Late Quaternary kinematics of and adjacent regions with special attention to the surroundings of the Pallatanga strike-slip fault 6central Ecuador) from topographic Panama. Tectonophysics 128, 21±46.